The effect of harmonized emissions on aerosol properties in global models – an AeroCom experiment

(1)

Chemistry

and Physics

The effect of harmonized emissions on aerosol properties in global

models – an AeroCom experiment

C. Textor1,19, M. Schulz1, S. Guibert1, S. Kinne2, Y. Balkanski1, S. Bauer3, T. Berntsen4, T. Berglen4, O. Boucher5,18, M. Chin16, F. Dentener6, T. Diehl17, J. Feichter2, D. Fillmore7,1, P. Ginoux9, S. Gong10, A. Grini4, J. Hendricks11, L. Horowitz9, P. Huang10, I. S. A. Isaksen4, T. Iversen4, S. Kloster2,6, D. Koch3, A. Kirkev˚ag4, J. E. Kristjansson4, M. Krol6,12, A. Lauer11, J. F. Lamarque7, X. Liu13,8, V. Montanaro14, G. Myhre4, J. E. Penner13, G. Pitari14, M. S. Reddy5,9, Ø. Seland4, P. Stier2,20, T. Takemura15, and X. Tie7

1_{Laboratoire des Sciences du Climat et de l’Environnement, Gif-sur-Yvette, France} 2_{Max-Planck-Institut f¨ur Meteorologie, Hamburg, Germany}

3_{Columbia University, GISS, New York, USA}

4_{University of Oslo, Department of Geosciences, Oslo, Norway}

5_{Laboratoire d’Optique Atmosph´erique, Universit´e des Sciences et Technologies de Lille, CNRS, Villeneuve d’Ascq, France} 6_{European Commision, Joint Research Centre, Institute for Environment and Sustainability, Climate Change Unit, Italy} 7_{NCAR, Boulder, Colorado, USA}

8_{Battelle, Pacific Northwest National Laboratory, Richland, USA}

9_{NOAA, Geophysical Fluid Dynamics Laboratory, Princeton, New Jersey, USA} 10_{ARQM Meteorological Service Canda, Toronto, Canada}

11_{DLR-Institut f¨ur Physik der Atmosph¨are, Oberpfaffenhofen, Germany}

12_{Institute for Marine and Atmospheric Research Utrecht (IMAU) Utrecht University, The Netherlands} 13_{University of Michigan, Ann Arbor, MI, USA}

14_{Universita degli Studi L’Aquila, Italy} 15_{Kyushu University, Fukuoka, Japan}

16_{NASA Goddard Space Flight Center, Greenbelt, MD, USA}

17_{Goddard Earth Sciences and Technology Center, University of Maryland Baltimore County, Baltimore, Maryland, USA} 18_{Hadley Centre, Met Office, Exeter, UK}

19_{Service d’A´eronomie, CNRS/UPMC/IPSL, Paris, France}

20_{Department of Environmental Science and Engineering, California Institute of Technology, Pasadena, USA}

Received: 18 December 2006 – Published in Atmos. Chem. Phys. Discuss.: 2 February 2007 Revised: 19 June 2007 – Accepted: 3 July 2007 – Published: 30 August 2007

Abstract. The effects of unified aerosol sources on global aerosol fields simulated by different models are examined in this paper. We compare results from two AeroCom ex-periments, one with different (ExpA) and one with unified emissions, injection heights, and particle sizes at the source (ExpB). Surprisingly, harmonization of aerosol sources has only a small impact on the simulated inter-model diversity of the global aerosol burden, and consequently global opti-cal properties, as the results are largely controlled by model-specific transport, removal, chemistry (leading to the forma-tion of secondary aerosols) and parameterizaforma-tions of aerosol microphysics (e.g., the split between deposition pathways) Correspondence to: C. Textor

(christiane.textor@lsce.ipsl.fr)

and to a lesser extent by the spatial and temporal distribu-tions of the (precursor) emissions.

The burdens of black carbon and especially sea salt be-come more coherent in ExpB only, because the large ExpA diversities for these two species were caused by a few out-liers. The experiment also showed that despite prescribing emission fluxes and size distributions, ambiguities in the im-plementation in individual models can lead to substantial dif-ferences.

(2)

are used to assess the consequences of specific air pollution abatement strategies.

1 Introduction

One of the largest uncertainties in assessing the human im-pact on climate is related to the role of aerosol and clouds (IPCC, 2007). The Aerosol interComparison project Aero-Com (http://nansen.ipsl.jussieu.fr/AEROCOM) attempts to advance the understanding of global aerosol and its impact on climate by performing a systematic analysis of the results of more than 16 global aerosol models including a compari-son with a large number of satellite and surface observations (Guibert et al., 20071; Kinne et al., 2006; Schulz et al., 2006; Textor et al., 2006). In these studies, it was found that signif-icant uncertainty in global modeling of spatial aerosol mass distributions is associated with aerosol processes.

The aerosol mass distributions depend on spatial and tem-poral distributions of emissions (of aerosols and precursors), on the ambient conditions (e.g., humidity or precipitation) and the transport in the atmosphere as described by the global transport models, as well as on the aerosol microphysical processes (e.g., water uptake or deposition, and chemistry for the production of secondary aerosols) as described by the im-plemented aerosol module. All these model components are inter-related. AeroCom focuses on the five most important aerosol components: dust (DU), sea salt (SS), sulfate (SO4),

black carbon (BC), and particulate organic matter (POM), and the sum of these components (AER).

In a first set of simulations (AeroCom ExpA, see Tex-tor et al., 2006, referred to as T2006 in the following) each model was run with emission data chosen by the individual participating groups. In these simulations, emission data dif-fered not only because of the use of diverse data sources, but even when referring to the same data source due to differ-ent implemdiffer-entation into the models (e.g., regridding, particle size-assumptions). In order to remove the impact of emission diversity on aerosol simulations a sensitivity experiment was performed (AeroCom ExpB) where unified global emission data sets for primary aerosol and aerosol precursors for the year 2000 were prescribed (Dentener et al., 2006).

In this study we compare simulated global mass distri-butions and underlying processes in AeroCom ExpA and ExpB. In the next two sections we summarize the

model-1_{Guibert, S., Schulz, M., Kinne, S., Textor, C., Balkanski, Y.,}

Bauer, S., Berntsen, T., Berglen, T., Boucher, O., Chin, M., Den-tener, F., Diehl, T., Feichter, H., Fillmore, D., Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Horowitz, L., Isaksen, I., Iversen, T., Kloster, S., Koch, D., Kirkev˚ag, A., Kristjansson, J. E., Krol, M., Lauer, A., Lamarque, J. F., Liu, X., Montanaro, V., Myhre, G., Pen-ner, J., Pitari, G., Reddy, S., Seland, O., Stier, P., Takemura, T., and Tie, X.: Comparison of lidar data with model results from the aerocom intercomparison project, in preparation, 2007.

setups and emissions. Then, changes in the diversity of sim-ulated global total aerosol mass distributions are presented and discussed in the context of spatial distributions and res-idence times of the different aerosol components. New ra-diative forcing estimates obtained from ExpB and an addi-tional experiment with unified sources for pre-industrial con-ditions are discussed in Schulz et al. (2006). Supplemen-tary maps and vertical profiles, and many other quantities (tagged FigS in the following) are provided in the electronic supplement to this paper (http://www.atmos-chem-phys.net/ 7/4489/2007/acp-7-4489-2007-supplement.pdf). Additional information can be found on the AeroCom web site (http: //nansen.ipsl.jussieu.fr/AEROCOM/data.html).

2 Model setup

A brief overview of the AeroCom models including a table linking model name abbreviations to the model versions ac-tually used can be found in Table 1, a comprehensive de-scription is given in T2006. The model configurations did not change between ExpA and ExpB, except for three mod-els: In the DLR model, coarse aerosols have been added only in ExpB. Larger changes have been made in KYU, where the interaction between aerosols and clouds has been included for ExpB, and carbonaceous aerosols (BC and POM) are treated externally, unlike the internal treatment in ExpA. In LOA, dry turbulent deposition is only considered in ExpA. In addition, deviations from the recommended AeroCom emis-sions occurred: In KYU and UIO GCM, sources of DU and SS remained those of ExpA. For the fine aerosols in KYU, only the emitted aerosol mass flux was matched, but size distributions have not been adapted. In ARQM, emis-sions have been modified for ExpB, but did not follow the ExpB recommendations. In MATCH, SS sources remained those of ExpA. Due to these deviations, all results of DLR, KYU, LOA, ARQM, as well as UIO GCM, and the SS and AER results of MATCH are discussed, but not included in the calculation of the model diversities. We also discard the results for all species in UIO GCM, although the emis-sions of BC, POM, and the sulfur species are consistent with the AeroCom ExpB recommandations. However, interac-tions among different aerosol types are taken into account in UIO GCM, because internal aerosol mixtures are con-sidered. Therefore, the results of all species are influenced by the non-AeroCom emissions of SS and DU. This is, by contrast, not the case in MATCH. Hence, we include GISS, LSCE, MATCH, MOZGN, UIO CTM, ULAQ and UMI in the statistics, MATCH is excluded from the calculations for SS and AER. The models that are excluded from the statis-tics are shaded in gray in the figures. Due to our sampling procedure, the statistics on ExpA reported in this paper do not entirely match the results reported in T2006.

(3)

Table 1. Description of the driver-models and general classification of aerosol modules. Models that follow the AeroCom ExpB

recommen-dations and are included in the calculations of the statistics are marked with $. The following abbreviations are used: accm: accumulation mode; bioburn: biomass burning; coag: coagulation; coam: coarse mode; cond: condensation; dyn: aerosol dynamics apart from sulfur cy-cle; ext: external; fix: prescribed; hetero: heterogeneous; insol: insoluble; int: internal; mix: mixture; modal M: modal scheme, mass mixing ratio as prognstic variable; modal MN: modal scheme, mass mixing ratio and number conc. as prognostic variables; nucl: nucleation; nuclm: nucleation mode; prog: prognostic; sigma-p: hybrid-sigma p; sol: soluble; supcoam: super coarse mode; thermodyn: thermodynamics.

Model Global model Horizont. Vertical References for Type of Number of Aerosol Aerosol References for

Resolution Resolution global model scheme bins or modes mixing dynamics aerosol module

(x y)(lon lat) (# of levels)

(type)

ARQM GCM 128×64 32 Zhang and McFarlane bin 12 int nucl, cond, coag Zhang et al. (2001)

Canadian 2.81◦_×2.81◦ sigma-p (1995b) all internally mixed thermodyn, Gong et al. (2003)

GCMIII cloud processing

DLR GCM 96×48 19 Roeckner et al. (1996) modal MN A:2, B:3 int nucl, cond, coag, Ackermann et al. (1998)

A: nucm+accm, B:+coam

GISS$ _GCM ₄₆_×₇₂ ₂₀ _{Schmidt et al. (2006)} _bin ₁₃ _ext _{aging BC POM,} _{Koch et al. (1999, 2006);}

modelE 5◦_×4◦ sigma 2 SS, 4 DU, 1 BC, 1 POM, hetero DU-SO4 Koch (2001); Bauer and

1 SO4, 4 DU/SO4 Koch (2005); Cakmur

et al. (2006) Koch and Hansen (2005) Miller et al. (2006)

KYU GCM 320×160 20 Numaguti et al. (1995); bin, 17 bins B: ext nucl, cond, coag, Takemura et al.

(SPRINT) CCSR/ NIES/ 1.1◦_×1.1◦ sigma Hasumi and Emori (2004) modal M 10 DU, 4 SS, 1 BC, thermodyn, (2000, 2002, 2005)

FRCGC for aerosol dyn∗ A; 1 BCPOM, 1 SO4 A: partly int B: cloud processing

SPRINTARS 5.7b (5 modes for aerosol dyn) for BCPOM

LOA GCM 96×72 19 Sadourny and Laval (1984); bin 16 ext aging BC POM Boucher and Anderson (1995);

LMDzT 3.3 3.75◦_×2.5◦ sigma Hourdin and Armengaud (1999) 2 DU, 11 SS, Boucher et al. (2002);

1 BC (sol+insol), Reddy and Boucher (2004);

1 POM (sol+insol), Guibert et al. (2005)

1 SO4

LSCE$ _GCM ₉₆_×₇₂ ₁₉ _{Sadourny and Laval (1984);} _{modal MN} ₅ _{ext mix of} _{aging BC POM} _{Claquin et al. (1998, 1999);}

LMDzT 3.3 3.75◦_×2.5◦ sigma Hourdin and Armengaud (1999) accm: sol+insol, int modes2 _{Guelle et al. (1998a, b, 2000)}

coam: sol+insol Smith and Harrison (1998)

supcoam: sol Balkanski et al. (2003)

Bauer et al. (2004)

Schulz et al. (2006)2

MATCH$ _CTM ₁₉₂_×₉₄ ₂₈ _{Zhang and McFarlane (1995a);} _bin ₈ _ext _{aging BC POM} _{Barth et al. (2000);}

MATCH v 4.2 1.9◦_×1.9◦ sigma-p Rasch et al. (1997) 4 DU, 1 SS,1 BC, Rasch et al. (2000, 2001);

Rasch and Kristjansson (1998) 1 POM, 1 SO4

MOZGN$ _CTM ₁₉₂_×₉₆ ₂₈ _{Brasseur et al. (1998)} _bin ₁₂ _ext _{aging BC POM} _{Tie et al. (2001, 2005)}

MOZART v2.5 1.9◦_×1.9◦ sigma-p Tie et al. (2001, 2005) 1 SO4, 1 POM

Horowitz et al. (2003) 1 BC, 5 DU, 4 SS

UIO CTM$ _CTM ₁₂₈_×₆₄ ₄₀ _{Berglen et al. (2004)} _bin ₂₅ _ext _{aging BC POM} _{Grini et al. (2002b, 2005)}

OsloCTM2 2.81◦_×2.81◦ sigma 8 DU, 8 SS, except bioburn Myhre et al. (2003)

4 BC, 4 POM, 1 SO4 Berglen et al. (2004)

Berntsen et al. (2006)

UIO GCM GCM 128×64 18 Hack (1994) modal, M/MN 12 modes 4 ext nucl, cond, coag, Iversen and Seland (2002)

CCM3.2 2.81◦_×2.81◦ sigma-p Kiehl et al. (1998) bin for aerosol dyn: 8 int: mixed thermodyn, Kirkev˚ag and Iversen (2002)

aerosol dyn 43 bins from +8 fix∗∗ cloud processing Kirkev˚ag et al. (2005)

8 int modes from 4 prog

DU+SS fix∗∗

ULAQ$ _CTM ₁₆_×₁₉ ₂₆ _{Pitari et al. (2002)} _bin ₄₁ _ext _{aging BC POM,} _{Pitari et al. (1993, 2002)}

ULAQ 22.5◦_×10◦ log-p 7 DU, 9 SS, 5 BC ext Koch, 2001

5 POM, 15 SO4 SO4microphysics

UMI CTM$ ₁₄₄_×₉₁ ₃₀ _{Schubert (1993);} _bin ₁₃ _Ext _none _{Liu and Penner 2002}

IMPACT 2.5◦_×2◦ sigma-p Rotman et al. (2004) 3 SO4, 1 POM,

1 BC, 4 DU, 4 SS

∗KYU describes the size distributions differently for transport and aerosol dynamics. 17 bins are considered for transport. For the aerosol dynamics, a modal approach is employed

(one mode per species, sigma fix).

∗∗UIO GCM describes the size distributions differently for transport and aerosol dynamics. 12 modes are considered for transport, 4 of them have prescribed size distributions,

are transported and not mixed with the other modes (external). The next 4 modes are also transported and only the shape of the distribution is constant (sigma fix). For the aerosol dynamics, these latter 4 modes are internally mixed with 8 prescribed modes, and fitted to 43 bins.

2_{Schulz, M., Balkanski, Y., Textor, C., Guibert, S., Generoso, S., Boucher, O., Breon, F.-M., Hauglustaine, D., and Hourdin, F.: The LMDzT-INCA global aerosol model and its}

evaluation with surface, lidar and satellite aerosol observations, in preparation, 2007.

components, SS and DU. This is caused by differences in the simulated size spectrum of the emitted particles, by differ-ences in the parameterizations of source strength as a func-tion of wind speed (and soil properties for DU), and by

dif-ferences in the wind fields themselves. Emissions of the “anthropogenic” species (SO4, BC, and POM) show better

(4)

Table 2. Statistics of models results for DU for the following parameters given in the rows: Emi: emissions, Load: burdens, Wet: wet

deposition fluxes, SedDry: total dry deposition fluxes, ResTime: tropospheric residence times, LoadAltF: mass fractions above 5 km height, LoadPolF: mass fractions in polar regions (south of 80◦S and north of 80◦N), WetofTot: split of removal pathways (mass fraction of wet removal in relation to total removal). The second column gives the unit, and the third the number of models included in the the calculations of the statistics. The next six columns show the means, medians and the model diversities (standard deviations) for experiments A and B.

DUST unit # mean median Stdev

ExpA ExpB ExpA ExpB ExpA ExpB

Emi Tg/a 7 1640,0 1630,0 1580,0 1670,0 30 4

Load Tg 7 22,7 21,3 21,3 20,3 21 21

Wet Tg/a 7 518,0 498,0 516,0 504,0 27 46

SedDry Tg/a 7 1130,0 1120,0 1040,0 1160,0 50 20

Life days 7 5,4 4,8 5,1 4,4 26 22

LoadAltF % 7 14,0 13,4 13,3 13,9 61 61

LoadPolF % 7 1,7 1,2 1,0 1,0 101 97

WetofTot % 7 34,9 30,8 36,6 30,3 43 47

Table 3. Statistics of models results for SS, see Table 2 for explanations.

SS unit # mean median Stdev

Emi Tg/a 5 8200,0 7720,0 3830,0 7740,0 100 3

Load Tg 6 7,9 12,7 6,5 12,0 69 31

Wet Tg/a 5 1320,0 1940,0 1090,0 2220,0 67 45

SedDry Tg/a 5 6880,0 5780,0 3260,0 5670,0 108 17

Life days 5 0,5 0,5 0,3 0,6 59 24

LoadAltF % 6 9,4 9,2 4,5 2,8 111 128

LoadPolF % 7 4,4 2,9 1,4 1,5 147 127

WetofTot % 5 21,0 25,3 21,0 28,2 58 45

Table 4. Statistics of models results for SO4, see Table 2 for explanations. Emi in the first row refers to the sum of emissions of chemical

production of SO₄.

SO4 unit # mean median Stdev

Emi Tg/a 5 183,0 164,0 205,0 176,0 26 23

Load Tg 7 2,2 2,1 2,2 2,3 25 25

Wet Tg/a 6 158,0 142,0 174,0 153,0 27 24

SedDry Tg/a 6 22,3 20,4 21,5 19,8 31 29

Life days 6 4,4 4,5 4,5 4,5 19 18

LoadAltF % 7 32,9 33,3 29,5 30,9 30 28

LoadPolF % 7 6,7 6,0 7,6 6,9 43 47

WetofTot % 6 86,9 86,9 87,5 87,0 5 5

have often been improved for certain species or emission types by the individual modelers, and their mix in each of the ExpA models is variable, see references in T2006.

The unified emission data used in ExpB have been re-compiled from various recently published inventories,

(5)

Table 5. Statistics of models results for BC, see Table 2 for explanations.

BC unit # mean median Stdev

Emi Tg/a 7 11,6 7,8 11,3 7,8 10 1

Load Tg 7 0,3 0,2 0,2 0,2 46 26

Wet Tg/a 6 9,3 6,2 9,4 6,2 17 11

SedDry Tg/a 6 2,4 1,6 2,6 1,5 33 38

Life days 7 8,1 7,6 7,2 7,2 41 26

LoadAltF % 7 21,3 22,8 17,4 17,5 54 53

LoadPolF % 7 5,1 3,9 6,1 4,3 66 67

WetofTot % 6 79,0 79,8 78,3 80,0 10 10

Table 6. Statistics of models results for POM, see Table 2 for explanations.

POM unit # mean median Stdev

Emi Tg/a 7 95,3 66,6 95,6 66,9 30 1

Load Tg 7 1,8 1,3 1,7 1,2 15 18

Wet Tg/a 6 81,8 53,0 79,5 52,1 33 11

SedDry Tg/a 6 19,9 13,0 19,9 13,9 25 38

Life days 7 7,4 6,9 7,1 6,7 26 18

LoadAltF % 7 21,4 22,5 17,5 19,7 59 54

LoadPolF % 7 3,9 3,3 4,6 3,9 63 65

WetofTot % 6 79,1 80,3 77,0 78,9 10 10

natural wild-land fires, and POM including secondary or-ganic aerosol. In addition, anthropogenic emissions from biomass burning and fossil fuel burning of SO2, POM and

BC are provided. The prescribed emission fields are gen-erated on a global 1◦_×₁◦_{spatial resolution, and a temporal}

resolution ranging from daily to annual. Injection heights for volcanic and wildfire emissions, and size distributions of the primary particulate emissions are prescribed. In this paper we focus on the emissions representative for present-day conditions. The models were nudged to (different) me-teorological data sets for the year 2000. Four models with-out nudging capability (General Circulation Models: ULAQ, UIO GCM, ARQM, and DLR), provided climatological av-erages from 5 years of simulation using the same emissions after a spin-up period of one year.

3 Results

3.1 Emissions

A comparison of the emissions in ExpA and ExpB shows that in most models the mass fluxes of “natural” aerosols (coarser sized SS and DU) are larger in the latter experiment (on av-erage by 87% and by 4%, respectively). The emissions of

carbonaceous species, BC and POM, are on average by 37% and by 27%, respectively, smaller. The total SO4 sources

decreased by 4%, see below for a discussion. For the model-average relative changes of the source mass fluxes see also Table 8. Model diversities for emissions and the all-model-averages and all-model medians for the annually and glob-ally averaged source fluxes are given in Tables 2–7 and in Figs. S1. Figures S2 show the annual mean zonally averaged emissions.

The implementation of the unified AeroCom sources in ExpB strongly reduced the diversity of global annual emis-sion mass fluxes when compared to ExpA. However, some differences in the emissions remained due to model-specific representations of the particle size distributions (bin schemes or modal schemes, and the number of modes or bins), or sim-ply by inaccurate implementation leading to small deviations from the AeroCom data. In addition, the initial degree of the mixing height, i.e., the impact of the sources on the vertical concentration profile, is governed by the model architecture (e.g., height of model levels and the emission scheme).

(6)

Table 7. Statistics of models results for AER, see Table 2 for explanations.

AER unit # mean median Stdev

Emi Tg/a 5 10100,0 9590,0 5930,0 9680,0 79 3

Load Tg 6 35,8 36,2 36,7 37,7 18 16

Wet Tg/a 5 2100,0 2610,0 2010,0 2940,0 50 42

SedDry Tg/a 5 8000,0 6960,0 5130,0 6870,0 89 17

Life days 5 1,9 1,3 1,7 1,4 61 16

LoadAltF % 6 15,3 14,5 12,4 11,2 54 62

LoadPolF % 7 2,9 2,3 2,0 1,8 94 96

WetofTot % 5 24,9 27,3 26,3 30,0 43 42

Table 8. Model average relative changes of parameters between ExpA and ExpB expressed as (ExpB-ExpA)/ExpA in [%] for the aerosol

species under consideration here (rows). The mean and the median changes are shown in second and third column for emissions, in the fourth and fifth column for loads, in the sixth and seventh column for residence times, and in the eighth and ninth column for the fractions of wet deposition in relation to total deposition, respectively.

Emi [Tg/a] Load [Tg] ResTime [days] WetofTot [%]

Mean Median Mean Median Mean Median Mean Median

DUST 3,8 3,0 17,5 −4,6 0,0 −9,9 −13,7 −0,9

SS 87,0 102,0 92,8 64,4 53,0 63,5 37,0 47,7

SO4 −4,1 0,2 9,1 1,1 1,5 5,7 2,1 0,2

BC −36,5 −31,9 −27,0 −32,8 6,4 4,4 2,6 1,5

POM −27,0 −30,7 −18,0 −24,6 5,9 8,0 2,4 2,2

AER 46,9 64,9 14,1 3,1 152,0 −22,3 35,5 24,2

ExpB to less than 5%, except for SO4. The diversityδof the

total SO4 source in ExpB (δ=23%) is almost as large as in

ExpA (δ=26%). SO4 originates predominantly (about 97%

on average in both experiments) from model-specific chem-ical production as sulfur-containing precursor gases (DMS and SO2)are oxidized. Direct emission of SO4 decreased

from ExpA to ExpB by 11% and that of the precursor gases SO2 and DMS by 11% and 50%, respectively. 79% (90%) of the secondary SO4stems from SO2, and 21% (10%) from

DMS oxidation in ExpA (ExpB). A comparison of the in-dividual processes involved in the sulphur cycle shows, that the diversity in SO4sources is due to differences in precursor

gas emissions, but differences in the dry deposition of these gases and the chemical production are much more important, see Figs. S3. The individual chemical production pathways in the liquid and gas phase show larger diversities than the total chemical production. This indicates compensation ef-fects. The production of SO4 from precursor gases leads

to the larger diversity of SO4 when compared to the other

aerosol components. Note however, that the statistics of the sulfur cycle is based on only three models, which delivered all quantities involved for both experiments.

3.2 Total mass

The changes in (global annual) masses for individual aerosol components between ExpA and B are generally consistent with those for emissions: models with increased emissions show larger mass and vice versa, for the model-average rela-tive changes of the total masses see also Table 8.

(7)

DUST SS SO4 BC POM AER

7 6 7 7 7 6 # models

21 69 25 46 15 18 Diversity

-100 -60 -20 20 60 100

[%]

130 102

(a)

DUST SS SO4 BC POM AER

7 6 7 7 7 6 # models

21 31 25 26 18 16 Diversity

-100 -60 -20 20 60 100

[%]

(b)

Fig. 1. Model diversities of the global, annual average aerosol

burden of the five aerosol species in (a) ExpA and (b) ExpB. The diversity ranges (_±Div =_±standard deviationδ) are indicated by gray boxes, the diversities are given in gray in the lower part of the plot for each species. The individual models’ deviations from the all-models-averages are plotted as pink lines (“data”), or as pink numbers if they are outside the scale of the plot. The all-models-averages are indicated by a black star (at 0%) and the medians by a black line (i.e., deviation of the median from the all-models-average). The numbers of models included in the calculation of this statistics are shown in blue below the x-axis.

3.3 Spatial distributions

Horizontal and vertical dispersal differed considerably among models that participated in AeroCom experiments (T2006), and harmonizing the sources in ExpB does not lead to significantly higher model agreement, see Figs. S5 and S6 electronic supplement for maps. In T2006, we defined two metrics describing the differences of the simulated spatial aerosols pattern. The degree of vertical dispersal was charac-terized by the mass fractions above 5 km height, where wet scavenging should become increasingly less significant due to the decrease in cloud precipitation efficiency. The degree of horizontal dispersal, i.e., meridional long-range transport, was described by the mass fractions in polar regions (defined

18

ARQM ARQMB DLR DLRB GISS GISSB KYU KYUB LOA LOAB LSCE LSCEB MATCH MATCHB MOZGN MOZGNB UIOCTM UIOCTMB UIOGCM UIOGCMB ULAQ ULAQB UMI UMIB

0 1 2 3 4 5 6 7 8 9 10 11 12

0

Fraction of total [%]

0 1 2 3 4 5 6 7 8 9 10 11 12

0

(a)

0 10 20 30 40 50 60

0

0 10 20 30 40 50 60

0

(b)

Fig. 2. (a) Global, annual average mass fractions in [%] of total

mass in polar regions (south of 80◦S and north of 80◦N) for the AeroCom models. (b) Global, annual average mass fractions in [%] of total mass above 5 km altitude for the AeroCom models. The dark gray shadings frame the range for each model. The models that are excluded from the statistics are shaded in light gray in the figures.

as south of 80◦S and north of 80◦N) because polar regions are far from most aerosol sources and not influenced by dif-ferences in local emissions. They thus reflect the long-range transport efficiency of the models.

(8)

11 11 15 11

0 1 2 3 4 5 6 7 8 9 10

[days]

(a)

105

218 138

1271 118

ARQM DLR GISS KYU LOA LSCE MATCH MOZGN UIOCTM UIOGCM ULAQ UMI -100 -80 -60 -40 -20 0 20 40 60 80 100

-100 -80 -60 -40 -20 0 20 40 60 80 100

(ExpB-ExpA)/ExpA [%]

(b)

Fig. 3. Tropospheric residence times in ExpA and ExpB in [days], (b) Relative changes between ExpA and B expressed as

(ExpB-ExpA)/ExpA in [%]. The models that are excluded from the statis-tics are shaded in light gray in the figures.

Fig. 2 shows, that spatial dispersal is more similar for any pair of simulations performed by an individual model than among all models using the same emissions in ExpB. (For additional layers at different heights (0–1, 1–2.5, 2.5–5, 5– 10,>10 km) see the electronic supplement in Fig. S9.) Thus, meridional and vertical dispersals seem to be determined by the model-specific combined effects of transport and the pa-rameterizations of internal aerosol processes.

3.4 Residence times

Another way of looking at model differences is the compar-ison of the (tropospheric) residence timeτ, which is defined as the ratio of global burden and emission. Figures 3 show residence times in ExpA and ExpB and the relative changes between both experiments. The residence time remained un-changed for DU, the relative increases ofτ in ExpB relative

to ExpA are on average 53% for SS, 2% for SO4, and 6%

for BC and POM, respectively, see also Table 8. The relative increase ofτ reached 152% for AER.

The variations ofτ between ExpA and ExpB are caused by the changes in spatial and temporal distribution of aerosol sources, that caused changes in the aerosol fields and thus in aerosol removal processes. The modification of the residence times of SO4can also be attributed to changes of pre-cursor

gas removal and of the conditions, under which its pre-cursor gases are oxidized to SO4, e.g., the coincidence of clouds and

SO2(See also the discussion in Sect. 3.1). Model diversities

for residence times (Tables 2–7, and Figs. S10) are somewhat smaller in ExpB than in ExpA, especially because some out-lier that occured in ExpA have been removed in ExpB.

The residence times depend on the simulated individual removal pathways. We examine these pathways and distin-guish between wet and dry deposition, where the latter com-prises turbulent deposition and sedimentation (see discussion in T2006). Fine aerosols (SO4, BC, and POM) are mainly

removed by wet deposition (on average about 80–90% by mass in both experiments, see also Tables 2–7). The split between the two removal pathways for fine aerosols is al-most exactly the same for al-most of the models, changes are smaller than 5% (see Fig. 4a for sulfate as an example for the fine fraction), and the diversity among models is similar in both experiments. (The figures for split between the removal path ways for the remaining species and the corresponding model diversities can be found in the electronic supplement Figs. S11 and S12.) Larger changes occur for LOA, KYU, and ARQM, but these models do not entirely fulfill the exper-iment requirements, see Sect. 2. The split between wet and dry removal is thus not sensitive to a change in emissions and associated assumptions on particles sizes. Since also the meteorological fields are equal in both experiments (at least for the CTMs) and thus the spatial distribution of clouds and precipitation, we can conclude that the changes in residence times for fine aerosols shown in Fig. 3 and Tables 2–7 are due to the changes in the spatial distribution of emissions (and de-position of precursor gases as well as chemical production in the case of SO4).

Another reason for the variations ofτ between the two ex-periments is the change in particles sizes (especially for the coarse aerosols SS and DU). The split of “fine” (diameter <1µm) and “coarse” (diameter>1µm) mode particles has been discussed in T2006. The mass fractions in the fine mode are 11% (7%), 18% (9%), 93% (93%), 99% (99%), and 99% (98%), for DU, SS, SO4, BC, and POM in ExpA (ExpB),

re-spectively. Figures S13 show the results in ExpA and ExpB and the corresponding model diversities. The models agree much better on the mass fractions of “anthropogenic” parti-cles in the fine mode (diversities ofδ=7% (δ=6%) for SO4in

(9)

in ExpA (ExpB)). This large disagreement can be explained by the model-specific representations of aerosol size distribu-tions. These lead to differences in the partinioning of the par-ticle sizes recommanded in ExpB on the model size classes, for a more detailed discussion see T2006. In addition, the dependence of the microphysical processes on particle size is not well known, and is differently parameterized in the models.

For the coarser aerosols (SS and DU), dry deposition is with about 65–80% of the removal mass fluxes the dominant process in both experiments (see Tables 2 and 3, and Fig. 4b as an example for SS). Changes between the two experiments exist even for those models which had been shown to have an equal pathway split for fine aerosol. Model diversity of the mass deposited by dry deposition decreased from ExpA to ExpB from 53% to 20% for DU, and from 107% to 19% for SS. These findings indicate the influence of harmonized size distributions in ExpB on the dry removal rates for SS, where at least the spatial distribution of the sources should be sim-ilar in both experiments. For DU, the larger model diversity of the sizes in ExpB is not transferred to the residence times. Therefore, in this case the spatial and temporal distribution of the sources seems to play a major role. However, for both coarse aerosol species, the split between the deposition path-ways is still rather model-specific and less dependent on the change in the sources.

The interested reader can find additional figures in the electronic supplement showing analyses as those performed in the T2006 paper for both experiments including effective sink rate coefficients in Figs. S14 and S15, and the split be-tween stratiform and convective wet removal in Figs. S16.

4 Discussion

The effects of aerosols on climate and air quality are highly uncertain, both in terms of their mechanisms (e.g., impor-tance of secondary organic aerosols, or indirect effects on clouds and precipitation) and in terms of their quantity. The assessment of aerosol effects necessitates a profound knowl-edge of the aerosol life cycle that can presently not be gained from observations alone. The application of numeri-cal models using high-quality inventories of aerosol precur-sor gas and primary aerosol emissions and parameterisations of aerosol processes in the atmosphere are required in order to evaluate coherent reduction strategies.

The realistic description of aerosol emissions in model-ing is very important. However, despite significant improve-ments of the inventories in recent years, large uncertainties remain. Crucial information about initial aerosols proper-ties, e.g., related to size distribution or composition, is still missing. In addition, the representation of such properties needs to be improved in the model parameterisations. In our paper, however, we do not examine the quality of emission inventories, but focus on differences among model results,

0 10 20 30 40 50 60 70 80 90 100

0

0 10 20 30 40 50 60 70 80 90 100

0

(a)

0 10 20 30 40 50 60 70 80 90 100

0

0 10 20 30 40 50 60 70 80 90 100

0

(b)

Fig. 4. Contribution of the individual removal processes to the total

sink mass flux (annually and globally averaged) for the AeroCom models for (a) SO4and (b) SS in [%]. The color code is given in the legend. Wet refers to wet deposition. If possible we show the individual dry sink rate coefficients (Tur: turbulent deposition, and Sed: sedimentation), otherwise the sum of the two processes (Dry = SedTur) is plotted. The models that are excluded from the statistics are shaded in light gray in the figures.

(10)

We show that even with (almost) identical emissions, the global species budgets and the simulated relative importance of the contributing processes remains largely unchanged in the individual models. Therefore, the overall model diver-sity did not considerably decrease in ExpB when compared to that of ExpA. This indicates, that the simulated fate of aerosols is to a large extent model-dependent and controlled by processes other than the diversity in emission input in cur-rent global aerosol models. Please note, that the model diver-sities calculated here are still based on a rather small number of models. Hence, the statistics are not be very robust, and the numbers for the diversities given in this study should be taken as the first estimation of the real diversity of aerosol process modelling on the global scale.

Usually, emission inventories are obtained from bottom-up techniques integrating all available information on the sources. Recently, top-down techniques have been applied in inverse studies using improved satellite information in com-bination with numerical models in order to infer strength and geographic distribution of the emissions (e.g., Zhang et al., 2005). Model studies have investigated the effect of modified aerosol emissions: Stier et al. (2006) demonstrated non-linear responses of global aerosol fields when modifying aerosol emissions in their simulations considering aerosol component interactions. De Meij et al. (2006) have evalu-ated the impact of differences in the EMEP and AEROCOM emission inventories on the simulated aerosol concentrations and optical depths in Europe, and demonstrated that seasonal variations in the emissions should be considered. All these studies show important effects when modifying the simulated aerosol emissions that had been neglected so far. Our re-sults indicate that the magnitude of the impact of emission changes on aerosol effects might depend on the individual model configuration. Recent studies have shown that the re-sults from ensemble simulations better represent the observa-tions than the individual models (e.g., McKeen et al., 2005; Pagowski et al., 2005; Vautard et al., 2006). This finding is also true for the AeroCom models. Therefore, we recom-mend to use an ensemble of models when aiming at a quan-titative assessement of the impacts from emission changes, until robust quality measures become available.

This paper shows that differences in aerosol sources alone cannot explain inter-model diversity. The question is, which are the possible causes of the remaining still large model di-versity obtained from the simulations in ExpB with unified sources? And which are the critical processes that cause the large differences between models? We have looked at dif-ferent processes and parameters and tried to group models according to their parameterisations or model architecture. We could, however, not determine any clear relation between the modeling results and the model structure. The processes involved in the aerosol life cycle are strongly inter-related. The importance of a specific process depends on that of all others, and it changes with the ambient conditions and with the aerosol type. For example, wet deposition is the main

sink process for highly soluble aerosols like sulfate, but in a region where wet scavenging is less important due to the lack of precipitation, dry deposition becomes the most important process.

More detailed investigations are neccessary that focus on specific regions, seasons, aerosol types and processes. In par-ticular, we expect clarifications from sensitivity studies com-paring tendencies of individual processes with constraints imposed on other processes. Tracer experiments are envis-aged to examine transport and aerosol dispersal patterns. In addition, we would like to point out, that the model diversity is not only caused by differences in aerosol process modeling but also influenced by the transport (advection and mixing) as well as the meteorological conditions (such as relative hu-midity, clouds and precipitation) provided by the global host models in which the aerosol modules are implemented. A re-cent paper by Liu et al. (2007) shows very large differences of the simulated aerosol fields when using different metero-logical datasets to drive the same aerosol module.

It is a major goal of AeroCom to compare model simula-tions against measurements. However, existing data mainly refer to surface concentrations or optical properties, because the observations of individual processes is very challenging. Kinne et al. (2006) have shown that although model simula-tions agree quite well with the observed total aerosol optical depth, the contributions from the individual components dif-fer considerably among models. A comparison of the sim-ulated optical depths from ExpA and ExpB in Figs. S17 in the electronic supplement shows, that the composition of the total optical depth is model-specific, and less dependent on aerosol sources, as for the parameters discussed above. Model diversity, see Figs. S17b and c, even slightly increases in ExpB with unified sources for the coarse aerosols (SS and DU), and hence for total aerosol.

5 Conclusions

In this paper, the effects of unified aerosol sources on the simulated aerosol fields have been examined. We compared the results of twelve models for two sets of simulations, one without any constraints on aerosol sources (ExpA), and one where mass fluxes, injection heights and particle sizes of emissions were prescribed (ExpB). Although the aerosol emissions’ diversity among models strongly decreased, we realize that it is not straightforward to implement prescribed aerosol (precursor) sources in exactly the same way into dif-ferent model configurations. Inconsistencies in the actually simulated source fluxes were caused by differences in the model architecture and the representation of the particle size distributions, or simply by inaccurate implementation.

(11)

and less depending on the properties of the aerosol sources. This shows that modeled aerosol life cycles depend to a large extent on model-specific differences for transport, removal, chemistry (e.g., formation of sulfate or secondary organ-ics) and parameterizations of aerosol microphysics, and to a lesser extent on the spatial and temporal distributions of the (precursor) emissions. These results indicate the need for a better understanding of aerosol life cycles at process level (including spatial dispersal and interaction with meteorologi-cal parameters and processes) in order to obtain more reliable results from global aerosol simulations. This is particularly important as such model results are used to assess the con-sequences of specific air pollution abatement strategies. We recommend to use an ensemble of model simulation, when assessing the impacts from emission changes, until robust quality measures become available.

The AeroCom initiative aims to better understand which processes are the main contributors to model diversity. The interdependence of the processes involved in the aerosol life cycle complicates this task. Detailed evaluation studies against measurements for different regions and different sea-sons and looking at specific processes are performed. Efforts are made to establish data test beds on a regional and sea-sonal basis that are sufficiently accurate to help evaluating specific processes in modeling. Additional simulations ded-icated to specific processes are envisaged, where individual parameterizations are tested within at least one global host model.

Acknowledgements. This work was supported by the European

Projects PHOENICS (Particles of Human Origin Extinguishing “natural” solar radiation In Climate Systems) and CREATE (Construction, use and delivery of an European aerosol database), and the French space agency CNES (Centre National des Etudes Spatiales). The authors would like to thank the Laboratoire des Sciences du Climat et de l’Environnement, Gif-sur-Yvette, France, and the Max-Planck-Institut f¨ur Meteorologie, Hamburg, Germany. The work of O. Boucher forms part of the Climate Prediction Programme of the UK Department for the Environment, Food and Rural Affairs (DEFRA) under contract PECD 7/12/37. We are grateful to the editor and for the suggestions received from the reviewers that helped to improve the quality of this paper.

Edited by: M. Kulmala

References

Ackermann, I. J., Hass, H., Ebel, M. M., Binkowski, F. S., and Shankar, U.: Modal Aerosol Dynamics for Europe: Develop-ment and first applications, Atmos. Environ., 32, 2981–2999, 1998.

Balkanski, Y., Schulz, M., Claquin, T., Moulin, C., and Ginoux, P.: Global emissions of mineral aerosol: formulation and valida-tion using satellite imagery, in: Emission of Atmospheric Trace Compounds, edited by: Granier, C., Artaxo, P., and Reeves, C. E., Kluwer, 253–282, 2003.

Bauer, S. E., Balkanski, Y., Schulz, M., Hauglustaine, D. A., and Dentener, F.: Global modeling of heterogeneous chemistry on mineral aerosol surfaces: Influence on tropospheric ozone chemistry and comparison to observations, J. Geophys. Res. A., 109(D2), D02304, doi:10.1029/2003JD003868, 2004.

Bauer, S. E. and Koch, D.: Impact of heterogeneous sulfate formation at mineral dust surfaces on aerosol loads and ra-diative forcing in the Goddard Institute for Space Studies general circulation model, J. Geophys. Res., 110, D17202, doi:10.1029/2005JD005870, 2005.

Barth, M. C., Rasch, P. J., Kiehl, J. T., Benkovitz, C. M., and Schwartz, S. E.: Sulfur chemistry in the NCAR CCM: Descrip-tion, evaluaDescrip-tion, features and sensitivity to aqueous chemistry, J. Geophys. Res., 106, 20 311–20 322, 2000.

Berglen, T. F., Berntsen, T. K., Isaksen, I. S. A., and Sundet, J. K.: A global model of the coupled sulfur/oxidant chemistry in the troposphere: The sulfur cycle, J. Geophys. Res., 109, D19310, doi:10.1029/2003JD003948, 2004.

Berntsen, T. K., Fuglestvedt, J. S., Myhre, G., Stordal, F., and Berglen, T. F.: Abatement of greenhouse gases: Does location matter?, Climatic Change, 74(4), 377–411, 2006.

Boucher, O. and Anderson, T. L.: GCM assessment of the sensi-tivity of direct climate forcing by anthropogenic sulfate aerosols to aerosol size and chemistry, J. Geophys. Res., 100, 26 117– 26 134, 1995.

Boucher, O., Pham, M., and Venkataraman, C.: Simulation of the atmospheric sulfur cycle in the Laboratoire de Meteorologie Dy-namique General Circulation Model, Model description, model evaluation, and global and European budgets, Note scientifique de l’IPSL, 23, 2002.

Cakmur, R. V., Miller, R. L., Perlwitz, J., Koch, D., Geogdzhayev, I. V., Ginoux, P., Tegen, I., and Zender, C. S.: Constraining the global dust emission and load by minimizing the difference between the model and observations, J. Geophys. Res., 111, D06207, doi:10.1029/2005JD005791, 2006.

Claquin, T., Schulz, M., and Balkanski, Y.: Modeling the mineral-ogy of atmospheric dust, J. Geophys. Res., 104, 22 243–22 256, 1999.

Claquin, T., Schulz, M., Balkanski, Y., and Boucher, O.: The in-fluence of mineral aerosol properties and column distribution on solar and infrared forcing by dust, Tellus B, 50, 491–505, 1998. Dentener, F., Kinne, S., Bond, T., Boucher, O., Cofala, J., Generos,

S., Ginoux, P., Gong, S., Hoelzemann, J. J., Ito, A., Marelli, L., Penner, J., Putaud, J.-P., Textor, C., Schulz, M., v. d. Werf, G. R., and Wilson, J.: Emissions of primary aerosol and precursor gases for the years 2000 and 1750 prescribed data-sets for AeroCom, Atmos. Chem. Phys., 6, 4321–4344, 2006,

http://www.atmos-chem-phys.net/6/4321/2006/.

de Meij, A., Krol, M., Dentener, F., Vignati, E., Cuvelier, C., and Thunis, P.: The sensitivity of aerosol in Europe to two different emission inventories and temporal distribution of emissions, At-mos. Chem. Phys., 6, 4287–4309, 2006,

(12)

doi:10.1029/2001JD002002, 2003.

Grini, A., Myhre, G., Zender, C. S., and Isaksen, I. S. A.: Model simulations of dust sources and transport in the global atmo-sphere: Effects of soil erodibility and wind speed variability, J. Geophys. Res., 110, D02205, doi:10.1029/2004JD005037, 2005. Grini, A., Zender, C. S., and Colarco, P. R.: Saltation Sandblasting behavior during mineral dust aerosol production, Geophys. Res. Lett., 29(18), 1868, doi:10.1029/2002GL015248, 2002b. Guelle, W., Balkanski, Y. J., Dibb, J. E., Schulz, M., and Dulac,

F.: Wet deposition in a global size-dependent aerosol transport model, 2. Influence of the scavenging scheme on 210Pb verti-cal profiles, surface concentrations, and deposition, J. Geophys. Res., 103, 28 875–28 891, 1998a.

Guelle, W., Balkanski, Y. J., Schulz, M., Dulac, F., and Monfray, P.: Wet deposition in a global size-dependent aerosol transport model, 1. Comparison of a 1 year 210Pb simulation with ground measurements, J. Geophys. Res., 103, 11 429–11 445, 1998b. Guelle, W., Balkanski, Y. J., Schulz, M., Marticorena, B.,

Berga-metti, G., Moulin, C., Arimoto, R., and Perry, K. D.: Model-ing the atmospheric distribution of mineral aerosol: Comparison with ground measurements and satellite observations for yearly and synoptic timescales over the North Atlantic, J. Geophys. Res., 105, 1997–2012, 2000.

Guibert, S., Matthias, V., Schulz, M., B¨osenberg, J., Eixmann, R., Mattis, I., Pappalardo, G., Perrone, M. R., Spinelli, N., and Vaughan, G.: The vertical distribution of aerosol over Europe – Synthesis of one year of EARLINET aerosol lidar measurements and aerosol transport modeling with LMDzT–INCA, Atmos. En-viron., 39, 2933–2943, 2005.

IPCC: Climate Change 2007 – The Physical Science Basis Work-ing Group I Contribution to the Fourth Assessment Report of the IPCC Corporate Author Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, 2007. Iversen, T. and Seland, O.: A scheme for process-tagged SO4

and BC aerosols in NCAR CCM3: Validation and sensitiv-ity to cloud processes, J. Geophys. Res. A., 107(D24), 4751, doi:10.1029/2001JD000885, 2002.

Kinne, S., Schulz, M., Textor, C., Guibert, S., Balkanski, Y., Bauer, S. E., Berntsen, T., Berglen, T., Boucher, O., Chin, M., Collins, W., Dentener, F., Diehl, T., Easter, R., Feichter, H., Fillmore, D., Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Herzog, M., Horowitz, L., Huang, P., Isaksen, I., Iversen, T., Koch, D., Kirkev˚ag, A., Kloster, S., Krol, M., Kristjansson, E., Lauer, A., Lamarque, J. F., Lesins, G., Liu, X., Lohmann, U., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Seland, O., Stier, P., Takemura, T., and Tie, X.: An AeroCom initial assessment – optical properties in aerosol component modules of global mod-els, Atmos. Chem. Phys., 6, 1815–1834, 2006,

Kirkev˚ag, A. and Iversen, T.: Global direct radiative forcing by process-parameterized aerosol optical properties, J. Geophys. Res., 107(D20), 4433, doi:10.1029/2001JD000886, 2002. Kirkev˚ag, A., Iversen, T., Seland, Ø., and Kristj´ansson, J. E.:

Re-vised schemes for aerosol optical parameters and cloud conden-sation nuclei in CCM-Oslo, in: Institute Report Series No. 28, Department of Geosciences, University of Oslo, Oslo, Norway, 2005.

Liu, X. and Penner, J. E.: Effect of Mt. Pinatubo H2SO4/H2O aerosol on ice nucleation in the upper troposphere using a global

chemistry and transport model (IMPACT), J. Geophys. Res., 107(D12), doi:10.1029/2001JD000455, 2002.

Liu, X., Penner, J. E., Das, B., Bergmann, D., Rodriguez, J. M., Strahan, S., Wang, M., and Feng, Y.: Uncer-tainties in global aerosol simulations: Assessment using three meteorological datasets, J. Geophys. Res., 112(D11212), doi:10.1029/2006JD008216, 2007.

McKeen, S., Wilczak, J., Grell, G., Djalalova, I., Peckham, S., Hsie, E.-Y., Gong, W., Bouchet, V., Menard, S., Moffet, R., McHenry, J., McQueen, J., Tang, Y., Carmichael, G. R., Pagowski, M., Chan, A., Dye, T., Frost, G., Lee, P., and Mathur, R.: Assessment of an ensemble of seven real-time ozone forecasts over eastern North America during the summer of 2004, J. Geophys. Res., 110(D21307), doi:10.1029/2005JD005858, 2005.

Miller, R. L., Cakmur, R. V., Perlwitz, J., Koch, D., Schmidt, G. A., Geogdzhayev, I. V., Ginoux, P., Prigent, C., and Tegen, I.: Min-eral dust aerosols in the NASA Goddard Institute for Space Sci-ences ModelE atmospheric general circulation model., J. Geo-phys. Res., 111(D06208), doi:10.1029/2005JD005796, 2006. Myhre, G., Berntsen, T. K., Haywood, J. M., Sundet, J. K., Holben,

B. N., Johnsrud, M., and Stordal, F.: Modelling the solar radia-tive impact of aerosols from biomass burning during the Southern African Regional Science Initiative (SAFARI-2000) experiment, J. Geophys. Res., 108, 8501, doi:10.1029/2002JD002313, 2003. Koch, D.: Transport and direct radiative forcing of carbonaceous and sulfate aerosols in the GISS GCM, J. Geophys. Res., 106, 20 311–20 322, 2001.

Koch, D. and Hansen, J.: Distant origins of Arctic Black Carbon: A GISS ModelE experiment, J. Geophys. Res., 110, D04204, doi:10.1029/2004JD005296, 2005.

Koch, D., Jacob, D., Tegen, I., Rind, D., and Chin, M.: Tro-pospheric sulfur simulation and sulfate direct radiative forcing in the Goddard Institute for Space Studies general circulation model, J. Geophys. Res. A., 104(D19), 23 799–23 822, 1999. Koch, D., Schmidt, G. A., and Field, C.: Sulfur, sea salt

and radionuclide aerosols in GISS, ModelE, J. Geophys. Res., 111(D06206), doi:10.1029/2004JD005550, 2006.

Pagowski, M., Grell, G. A., McKeen, S. A., D´ev´enyi, D., Wilczak, J. M., Bouchet, V., Gong, W., McHenry, J., Peckham, S., Mc-Queen, J., Moffet, R., and Tang, Y.: A simple method to im-prove ensemble-based ozone forecasts, Geophys. Res. Lett., 32, doi:10.1029/2004GL022305, 2005.

Pitari, G., Mancini, E., Rizi, V., and Shindell, D. T.: Impact of future climate and emissions changes on stratospheric aerosols and ozone, J. Atmos. Sci., 59, 414–440, 2002.

Pitari, G., Rizi, V., Ricciardulli, L., and Visconti, G.: High-speed civil transport impact: Role of sulfate, nitric acid trihydrate, and ice aerosol studied with a two-dimensional model including aerosol physics, J. Geophys. Res., 98, 23 141–23 164, 1993. Rasch, P. J., Collins, W. D., and Eaton, B. E.: Understanding the

INDOEX aerosol distributions with an aerosol assimilation, J. Geophys. Res., 106, 7337–7355, 2001.

(13)

Schwartz, S., Walcek, C., and Williamson, D. L.: A comparison of scavenging and deposition processes in global models: results from the WCRP Cambridge Workshop of 1995, Tellus B, 52, 1025–1056, 2000.

Reddy, M. S. and Boucher, O.: Global carbonaceous aerosol trans-port and assessment of radiative effects in the LMDZ GCM, J. Geophys. Res., 109(D14), D14202, doi:10.1029/2003JD004048, 2004.

Schulz, M., Textor, C., Kinne, S., Balkanski, Y., Bauer, S. E., Berntsen, T., Berglen, T., Boucher, O., Dentener, F., Grini, A., Guibert, S., Iversen, T., Koch, D., Kirkev˚ag, A., Liu, X., Mon-tanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Seland, Ø., Stier, P., and Takemura, T.: Radiative forcing by aerosols as derived from the AeroCom present-day and pre-industrial simu-lations, Atmos. Chem. Phys., 6, 5225–5246, 2006,

Stier, P., Feichter, J., Kloster, S., Vignati, E., and Wilson, J.: Emission-Induced Nonlinearities in the Global Aerosol System: Results from the ECHAM5-HAM Aerosol-Climate Model, J. Climate, 19, 3845–3862, 2006.

Takemura, T., Nozawa, T. T., Emori, S., Nakajima, T. Y., and Naka-jima, T.: Simulation of climate response to aerosol direct and in-direct effects with aerosol transport-radiation model, J. Geophys. Res., 110, D02202, doi:10.1029/2004JD005029, 2005.

Takemura, T., Nakajima, T., Dubovik, O., Holben, B. N., and Kinne, S.: Single-scattering albedo and radiative forcing of var-ious aerosol species with a global three-dimensional model, J. Climate, 15(4), 333–352, 2002.

Takemura, T., Okamoto, H., Maruyama, Y., Numaguti, A., Hig-urashi, A., and Nakajima, T.: Global three-dimensional simula-tion of aerosoloptical thickness distribusimula-tion of various origins, J. Geophys. Res., 105, 17 853–17 873, 2000.

Textor, C., Schulz, M., Kinne, S., Guibert, S., Balkanski, Y., Bauer, S. E., Berntsen, T., Berglen, T., Boucher, O., Chin, M., Dentener, F., Diehl, T., Easter, R., Feichter, H., Fillmore, D., Ghan, S., Gi-noux, P., Gong, S., Grini, A., Hendricks, J., Horowitz, L., Huang, P., Isaksen, I., Iversen, T., Kirkev˚ag, A., Kloster, S., Koch, D., Kristjansson, E., Krol, M., Lauer, A., Lamarque, J. F., Liu, X., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Se-land, Ø., Stier, P., Takemura, T., and Tie, X.: Analysis and quan-tification of the diversities of aerosol life cycles within AeroCom, Atmos. Chem. Phys., 6, 1777–1813, 2006,

Tie, X., Brasseur, G., Emmons, L., Horowitz, L., and Kinnison, D.: Effects of aerosols on tropospheric oxidants: A global model study, J. Geophys. Res., 106, 22 931–22 964, 2001.

Tie, X. X., Madronich, S., Walters, S., Edwards, D. P., Ginoux, P., Mahowald, N., Zhang, R. Y., Lou, C., and Brasseur, G.: Assess-ment of the global impact of aerosols on tropospheric oxidants, J. Geophys. Res. A., 110, D03204, doi:10.1029/2004JD005359, 2005.

Vautard, R., van Loon, M., Schaap, M., Bergstrom, R., Bessagnet, B., Brandt, J., Builtjes, P. J. H., Christensen, J. H., Cuvelier, K., Graff, A., Jonson, J. E., Krol, M., Langner, J., Roberts, P., Rouil, L., Stern, R., Tarrason, L., Thunis, P., Vignati, E., White, L., and Wind, P.: Is regional air quality model diversity representa-tive of uncertainty for ozone simulation?, Geophys. Res. Lett., 33(L24818), doi:10.1029/2006GL027610, 2006.

Zhang, L., Gong, S.-L., Padro, J., and Barrie, L.: A Size-segregated Particle Dry Deposition Scheme for an Atmospheric Aerosol Module, Atmos. Environ., 35(3), 549–560, 2001.